- This event has passed.
Armis 2 & Lighthouse Summer 2021 Maintenance
August 10, 2021 @ 6:00 am - August 11, 2021 @ 5:00 pm
Due to maintenance, the Armis 2 and Lighthouse clusters and their storage systems (/home and /scratch) will be unavailable Tuesday, August 10, 2021, 6 a.m. – Wednesday, August 11, 2021, 5 p.m.
Copy any files you might need during the maintenance window to your local drive using Globus File Transfer.
Contact firstname.lastname@example.org if you have any questions. More details and updates are available here.
- CentOS 7.9
- Slurm 20.11
- NVIDIA driver updates
- OFED updates
- Policy change: Slurm jobs will no longer automatically re-queue after a hardware failure. Slurm jobs can enable re-queueing in their slurm submission script with the
- New default versions for the following software packages:
|Module Name||Old Default Version||New Default Version|
Due to a change in the cluster configuration that affects process placement for MPI jobs, a number of module files have been updated to provide a consistent user experience. The modifications include setting new Slurm variables and adding a wrapper script that will be called with some Intel MPI (IMPI) modules when using
mpirun on a login node.
Jobs that run the following software/commands will likely be affected:
mpirun, matlab, mpiexec.hydra, CESM (with IMPI), LSDYNA, Abaqus, some ANSYS apps, or anything else that uses IMPI.
If you notice that your jobs using any of the indicated commands are running more slowly, please let us know by submitting a ticket to email@example.com.